Hangcheck is our back up in case the GPU or the driver gets stuck. It
detects when the GPU is not making any progress and issues a GPU reset.
However, if the driver is failing to make any progress, we can get
ourselves into a situation where we continually try resetting the GPU to
no avail. Employ a second timeout such that if we continue to see the
same seqno (the stalled engine has made no progress at all) over the
course of several hangchecks, declare the driver wedged and attempt to
start afresh.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180602104853.17140-1-chris@chris-wilson.co.uk
Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
In the unlikely case where we have failed to keep submitting to the GPU,
we end up with the ELSP queue empty but a pending queue of requests.
Here, we skip the per-engine reset as there is no guilty request, but in
doing so we also skip the engine restart leaving ourselves with a
permanently hung engine. A quick way to recover is by moving the tasklet
kick to execlists_reset_finish() (from init_hw). We still emit the error
on hanging, so the error is not lost but we should be able to recover.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Cc: Mika Kuoppala <mika.kuoppala@intel.com>
Cc: Tvrtko Ursulin <tvrtko.ursulin@intel.com>
Cc: Michel Thierry <michel.thierry@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180604073441.6737-2-chris@chris-wilson.co.uk
Reviewed-by: Mika Kuoppala <mika.kuoppala@intel.com>
With an old (4.7.3 on 32bit) gcc, it emits a warning for
In file included from drivers/gpu/drm/i915/i915_request.c:1425:0:
drivers/gpu/drm/i915/selftests/i915_request.c: In function ‘live_nop_request’:
drivers/gpu/drm/i915/selftests/i915_request.c:380:21: error: ‘request’ may be used uninitialized in this function [-Werror=maybe-uninitialized]
Silence it by just setting it to NULL on initialisation.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180614124923.18071-1-chris@chris-wilson.co.uk
While Bspec doesn't list a specific sequence for turning off the DP port
on g4x we are getting an underrun if the port is disabled in the
.disable() hook. Looks like the pipe stops when the port stops, and by
that time the plane disable may not have completed yet. Also the plane(s)
seem to end up in some wonky state when this happens as they also signal
another underrun immediately after we turn them back on during the next
enable sequence.
We could add a vblank wait in .disable() to avoid wedging the planes,
but I assume we're still tripping up the pipe in some way. So it seems
better to me to just follow the ILK+ sequence and turn off the DP port
in .post_disable() instead. This sequence doesn't seem to suffer from
this problem. Could be it was always the intended sequence for DP and
the gen4 bspec was just never updated to include it.
Originally we used the bad sequence even on ilk+, but I changed that
in commit 08aff3fe26 ("drm/i915: Move DP port disable to post_disable
for pch platforms") as it was causing issues on those platforms as well.
I left out g4x then only because I didn't have the hardware to test it.
Now that I do it's fairly clear that the ilk+ sequence is also the
right choice for g4x.
v2: Fix whitespace fail (Jani)
Mention the ilk+ commit (Jani)
Cc: stable@vger.kernel.org
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180613160553.11664-2-ville.syrjala@linux.intel.com
Reviewed-by: Jani Nikula <jani.nikula@intel.com>
On i965/g4x IIR is edge triggered. So in order for IIR to notice that
there is still a pending interrupt we have to force and edge in ISR.
For the ISR/IIR pipe event bits we can do that by temporarily
clearing all the PIPESTAT enable bits when we ack the status bits.
This will force the ISR pipe event bit low, and it can then go back
high when we restore the PIPESTAT enable bits.
This avoids the following race:
1. stat = read(PIPESTAT)
2. an enabled PIPESTAT status bit goes high
3. write(PIPESTAT, enable|stat);
4. write(IIR, PIPE_EVENT)
The end result is IIR==0 and ISR!=0. This can lead to nasty
vblank wait/flip_done timeouts if another interrupt source
doesn't trick us into looking at the PIPESTAT status bits despite
the IIR PIPE_EVENT bit being low.
Before i965 IIR was level triggered so this problem can't actually
happen there. And curiously VLV/CHV went back to the level triggered
scheme as well. But for simplicity we'll use the same i965/g4x
compatible code for all platforms.
Cc: stable@vger.kernel.org
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=106033
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=105225
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=106030
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180611200258.27121-1-ville.syrjala@linux.intel.com
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
The immediate enabling was actually not an issue for the
HW perspective for core platforms that have HW tracking.
HW will wait few identical idle frames before transitioning
to actual psr active anyways.
Now that we removed VLV/CHV out of the picture completely
we can safely remove any delays.
Note that this patch also remove the delayed activation
on HSW and BDW introduced by commit 'd0ac896a477d
("drm/i915: Delay first PSR activation.")'. This was
introduced to fix a blank screen on VLV/CHV and also
masked some frozen screens on other core platforms.
Probably the same that we are now properly hunting and fixing.
v2:(DK): Remove unnecessary WARN_ONs and make some other
VLV | CHV more readable.
v3: Do it regardless the timer rework.
v4: (DK/CI): Add VLV || CHV check on cancel work at psr_disable.
v5: Kill remaining items and fully rework activation functions.
v6: Rebase on top of VLV/CHV clean-up and keep the reactivation
on a regular non-delayed work to avoid extra delays on exit
calls and allow us to add few more safety checks before
real activation.
Cc: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com>
Signed-off-by: Rodrigo Vivi <rodrigo.vivi@intel.com>
Reviewed-by: José Roberto de Souza <jose.souza@intel.com>
Reviewed-by: Dhinakaran Pandiyan <dhinakaran.pandiyan@intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180613192600.3955-1-rodrigo.vivi@intel.com
Fix the issue that SCLK&MCLK can't be set higher than dpm7 when
OD is enabled in SMU7.
v2: fix warning (Alex)
Signed-off-by: Kenneth Feng <kenneth.feng@amd.com>
Acked-by: Rex Zhu<rezhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
This patch makes RC_CORE to be selected with this driver.
sil_sii8620 driver calls remote controller interfaces directly
so RC_CORE should be enabled mandatorily.
And some boards not using remote controller device don't really
need to know that RC_CORE config should be enabled to use sil_sii8620
driver only for HDMI.
Changelog v2:
- select INPUT because compiling will fail without INPUT.
Signed-off-by: Inki Dae <inki.dae@samsung.com>
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1527154379-31886-1-git-send-email-inki.dae@samsung.com
For symmetry, simplicity and ensuring the request is always truly idle
upon its completion, always emit the closing flush prior to emitting the
request breadcrumb. Previously, we would only emit the flush if we had
started a user batch, but this just leaves all the other paths open to
speculation (do they affect the GPU caches or not?) With mm switching, a
key requirement is that the GPU is flushed and invalidated before hand,
so for absolute safety, we want that closing flush be mandatory.
Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>
Reviewed-by: Joonas Lahtinen <joonas.lahtinen@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180612105135.4459-1-chris@chris-wilson.co.uk
Gfxoff is already enabled in amdgpu_device_ip_set_powergating_state.
So, no need to enable it again in pp_late_init.
Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
WARN_ON possible buffer overflow and avoid unnecessary dereference.
v2: change BUG_ON to WARN_ON
Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
As hw required, soc clock must large than mclk, So we set max soc
clock to OD Max Memory clk.
But on workstation, vbios do not support OD feature, the OD max memory
clock is equal to 0. In this case, driver can support underclocking.
and set od max memory clock to the value in highest memory dpm level.
So the od max memory clock should be less than highest soc clock.
and driver should not change the soc clock.
caused by commit ca57b9b0a156
("drm/amd/pp: Allow underclocking when od table is empty in vbios")
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Fixes stale buffer object (bo) usage for cursor plane
Cursor plane's bo operations are handled in DC code.
Currently, atomic_commit() does not handle bo operations
for cursor plane, as a result the bo assigned for cursor
plane in dm_plane_helper_prepare_fb() is not coherent
with the updates to the same made in dc code.This mismatch
leads to "bo" corruption and hence crashes during S3 entry.
This patch cleans up the code which was added as a hack
for 4.9 version only.
Reviewed-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Signed-off-by: Pratik Vishwakarma <Pratik.Vishwakarma@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The current use of result is or'ing in values and checking for
a non-zero result, however, result is not initialized to zero
so it potentially contains garbage to start with. Fix this by
initializing it to the first return from the call to
vega10_program_didt_config_registers.
Detected by cppcheck:
"(error) Uninitialized variable: result"
Fixes: 9b7b8154cd ("drm/amd/powerplay: added didt support for vega10")
Signed-off-by: Colin Ian King <colin.king@canonical.com>
Acked-by: Huang Rui <ray.huang@amd.com>
[Fix the subject as Colin's comment]
Signed-off-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
The clocks should be adjusted after display configuration changed.
Otherwise, the socclk and memclk may be forced on an unnecessary higher
level.
Signed-off-by: Evan Quan <evan.quan@amd.com>
Reviewed-by: Rex Zhu <Rex.Zhu@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
So, unfortunately I recently made the discovery that in the upstream
kernel, the only reason that amdgpu is not currently suffering from
issues with runtime PM putting the GPU into suspend while it's driving
displays is due to the fact that on most prime systems, we have sound
devices associated with the GPU that hold their own runtime PM ref for
the GPU.
What this means however, is that in the event that there isn't any kind
of sound device active (which can easily be reproduced by building a
kernel with sound drivers disabled), the GPU will fall asleep even when
there's displays active. This appears to be in part due to the fact that
amdgpu has not actually ever relied on it's rpm_idle() function to be
the only thing keeping it running, and normally grabs it's own power
references whenever there are displays active (as can be seen with the
original pre-DC codepath in amdgpu_display_crtc_set_config() in
amdgpu_display.c). This means it's very likely that this bug was
introduced during the switch over the DC.
So to fix this, we start grabbing runtime PM references every time we
enable a previously disabled CRTC in atomic_commit_tail(). This appears
to be the correct solution, as it matches up with what i915 does in
i915/intel_runtime_pm.c.
The one sideaffect of this is that we ignore the variable that the
pre-DC code used to use for tracking when it needed runtime PM refs,
adev->have_disp_power_ref. This is mainly because there's no way for a
driver to tell whether or not all of it's CRTCs are enabled or disabled
when we've begun committing an atomic state, as there may be CRTC
commits happening in parallel that aren't contained within the atomic
state being committed. So, it's safer to just get/put a reference for
each CRTC being enabled or disabled in the new atomic state.
Signed-off-by: Lyude Paul <lyude@redhat.com>
Acked-by: Christian König <christian.koenig@amd.com>.
Reviewed-by: Harry Wentland <harry.wentland@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
Cc: stable@vger.kernel.org
Patch 9667849bbb: "drm/amd/powerplay: add control gfxoff enabling in late
init" from Mar 13, 2018, leads to the following static checker warning:
drivers/gpu/drm/amd/amdgpu/../powerplay/amd_powerplay.c:194
pp_late_init()
error: we previously assumed 'hwmgr' could be null (see line 185)
drivers/gpu/drm/amd/amdgpu/../powerplay/amd_powerplay.c
This patch fixes the warning to add hwmgr checking.
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Evan Quan <evan.quan@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
After defer the execution of clockgating enabling, at that time, gfx already
enter into "off" state. Howerver, clockgating enabling will use MMIO to access
the gfx registers, then get the gfx hung.
So here we should move the gfx powergating and gfxoff enabling behavior at the
end of initialization behind clockgating.
Signed-off-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang@amd.com>
Acked-by: Christian König <christian.koenig@amd.com>
Cc: Felix Kuehling <Felix.Kuehling@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
The sos ucode version will be changed to align with the value of
mmMP0_SMN_C2PMSG_58. Then we add a checking for this. Meanwhile, we have to be
compatibility backwards. So it adds serveral recent legacy versions as the white
list for the version checking.
Signed-off-by: Huang Rui <ray.huang@amd.com>
Acked-by: Alex Deucher <alexander.deucher@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
When encountering a connector with the scaling mode property both
intel and modesetting ddxs sometimes add tons of DBLSCAN modes
to the output's mode list. The idea presumably being that since the
output will be going through the panel fitter anyway we can pretend
to use any kind of mode.
Sadly that means we can't reject user modes with the DBLSCAN flag
until we know whether we're going to be using the panel's native
mode or the user mode directly. Doing otherwise means X clients using
xf86vidmode/xrandr will get a protocol error (and often self
terminate as a result) when the kernel refuses to use the requested
mode with the DBLSCAN flag.
To undo the regression we'll move the DBLSCAN checks into the
connector->mode_valid() and encoder->compute_config() hooks.
Cc: stable@vger.kernel.org
Cc: Vito Caputo <vcaputo@pengaru.com>
Reported-by: Vito Caputo <vcaputo@pengaru.com>
Fixes: e995ca0b81 ("drm/i915: Provide a device level .mode_valid() hook")
References: https://lkml.org/lkml/2018/5/21/715
Signed-off-by: Ville Syrjälä <ville.syrjala@linux.intel.com>
Link: https://patchwork.freedesktop.org/patch/msgid/20180524125403.23445-1-ville.syrjala@linux.intel.com
Reviewed-by: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=106804
Tested-by: Arkadiusz Miskiewicz <arekm@maven.pl>
MHL bridge is usually connected to TV via MHL dongle. Currently plugging
HDMI cable to dongle is handled improperly.
Fix it by splitting connecting of a dongle and a HDMI cable. The driver
should now handle unplugging a sink from a dongle and plugging a
different sink with new edid.
Tested on MHL1, MHL2 and MHL3 using various vendors' dongles both in
DVI and HDMI mode.
Signed-off-by: Maciej Purski <m.purski@samsung.com>
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1516705996-8928-1-git-send-email-m.purski@samsung.com
The vendor code waits for infoframe to detect video mode set by source.
We do not need to follow this pattern, because video mode information is
provided by drm core. As a result most of the infoframe handling
code can be removed.
Start transmission immediately after detecting stream on HDMI lines
in irq_scdt() function without waiting for infoframe interrupt.
Signed-off-by: Maciej Purski <m.purski@samsung.com>
Signed-off-by: Andrzej Hajda <a.hajda@samsung.com>
Link: https://patchwork.freedesktop.org/patch/msgid/1511956130-24482-1-git-send-email-m.purski@samsung.com
Enable KVMGT for BXT.
is_supported_device() acting as the gatekeeper of GVT-g init.
If all supported platforms share the same configurations for some
specific feature, platform check will rely on this check only.
Signed-off-by: Colin Xu <colin.xu@intel.com>
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
Leverage most SKL/KBL mmio init info and add different mmio to
BXT specific function init_bxt_mmio_info().
Signed-off-by: Colin Xu <colin.xu@intel.com>
Signed-off-by: Zhenyu Wang <zhenyuw@linux.intel.com>
